Don Casey - Dragged Aboard Dragged Aboard by Don Casey
A Cruising Guide for the Reluctant Mate


      

Other books by Don Casey
| Home | Mailing Lists | Bookstore | Weather | Tide Predictions | Bowditch |

Re: ExcelNav.ZIP Question

From: (no name) (no email)
Date: Tue Apr 18 2006 - 11:25:15 EDT

  • Next message: George Huxtable: "Re: Error of Perpendicularity"

    On Saturday, April 08, 2006 10:20 AM, Robert Gainer said:
    > Not being a person that can program an Excel spreadsheet, I am not even
    > sure how to ask the question. In my copy of Excel, using the dialog box
    > for formatting cells I see you can define an input field as "custom" and
    > have input in combinations of time and number. That is to say, enter a
    > number + minuets, and seconds (##0 MM SS) or just HHH MM SS using that
    > custom mask. Why can't the spreadsheet you wrote use time as a stand-in
    > for degrees, minuets and seconds of an angle. The spreadsheet understands
    > that a minuet and a second have 60 integers in them so it would seem like
    > you can trick it into using that input mask to make a user friendly front
    > end. If I understand it correctly, you can unstring the data to use in
    > calculations later in the spreadsheet so it looks doable in the company
    > of a sufficient amount of ignorance on my part.

    Interesting idea, but I'm not sure it really simplifies the sheets.

    As they stand, the sheets are built to accept angles as decimal degrees.
    Thus, 12° 34.5' is entered as +12+34.5/60. An adjacent cell displays this
    formatted as degrees, minutes and tenths for clarity. Calculations are
    carried out on the decimal degrees; trig functions are wrapped in the
    DEGREES() and RADIANS() functions to satisfy their appetite for radians.
    This makes the spreadsheets look similar to paper work forms (and easier to
    debug). This is actually why I wrote these spreadsheets to replace the C
    programs (derived from earlier Fortran programs) that I had used for years:
    to have a visible, step-by-step solution rather than a black box that spat
    out answers.

    Using Excel's native HH:MM:SS format has two problems.

    First, the resulting numbers, which Excel thinks are time, are stored as
    fractions. Excel's date-time format encodes dates as days since 1-Jan-1900
    and time of day as fractional days. Having angles expressed as fractions
    [0,1] is neither as intuitive as decimal degrees [0,360], nor as
    computationally convenient as radians [0,2pi].

    Second, there's a gotcha in the way Excel interprets time entries. 12:34:56
    is unambiguously interpreted as hours:minutes:seconds and resolves to
    0.524259 (12/24 + 34/1440 + 56/86400). Likewise, 12:34 is interpreted as
    hours:minutes and resolves to 0.523611 (12/24 + 34/1440). BUT 12:34.5, which
    to a navigator would imply 12° 34.5', is interpreted by Excel as
    minutes:seconds and resolves to 0.008733 (12/1400 + 34.5/86400). Thus
    XX:YY[.Z] is interpreted differently depending on whether a fraction is
    present. One could probably kluge around this, but you'd be steering into
    danger.

     -- Peter

    -----Original Message-----
    From: Navigation Mailing List [mailto:]
    On Behalf Of Robert Gainer
    Sent: Saturday, April 08, 2006 10:20 AM
    To:
    Subject: ExcelNav.ZIP Question

    Peter, and others
    Not being a person that can program an Excel spreadsheet, I am not even sure
    how to ask the question. In my copy of Excel, using the dialog box for
    formatting cells I see you can define an input field as "custom" and have
    input in combinations of time and number. That is to say, enter a number +
    minuets, and seconds (##0 MM SS) or just HHH MM SS using that custom mask.
    Why can't the spreadsheet you wrote use time as a stand-in for degrees,
    minuets and seconds of an angle. The spreadsheet understands that a minuet
    and a second have 60 integers in them so it would seem like you can trick it
    into using that input mask to make a user friendly front end. If I
    understand it correctly, you can unstring the data to use in calculations
    later in the spreadsheet so it looks doable in the company of a sufficient
    amount of ignorance on my part.
    Thanks,
    Robert Gainer


  • Next message: George Huxtable: "Re: Error of Perpendicularity"



    | Home | Mailing Lists | Bookstore | Weather | Tide Predictions | Bowditch | Trawlerworld |